KR101420178B1 - Method of enabling a packet loss measurement in a packet transport network - Google Patents

Abstract

The proposed method is to enable packet loss measurement in the network. Data packets comprising the same source address and the same destination address are received through a group of user ports. Link aggregation of the user ports group is performed. At the network port, the received data packets are sent to the far-end network edge node. The number of transmitted data packets is determined and the number of internally lost data packets for each of the user ports is determined. At the network port, the modified number of transmitted data packets is indicated to the far-end network edge node as the sum of the determined number of transmitted data packets and the total number of internally lost data packets.

Description

[0001] METHOD OF ENABLING A PACKET LOSS MEASUREMENT IN A PACKET TRANSPORT NETWORK [0002]

The present invention relates to a method of packet loss measurement in a network.

When receiving a plurality of data packets at a network edge node and transmitting the data packets over a network to the far end network edge node, the number of lost data packets during this transmission is monitored by performance monitoring ). ≪ / RTI > A protocol for packet loss measurement is already known as proposed in the publication " ITU-T Recommendation Y.1731 - OAM functions and mechanisms for Ethernet based networks, (02/2008) ", simply called Y.1731 , Which provides a method of packet loss measurement between two packet processors in which the protocol is implemented. Thus, if a packet processor is provided that runs a protocol on each of the network ports, the two network edge nodes can determine the number of data packets lost during transmission between their network ports. Therefore, the number of lost data packets between two clients, each connected to one of the network edge nodes, can be determined.

The present invention aims at improving the above-described known method of packet loss measurement.

The proposed method is to enable packet loss measurement in the network. The method is comprised of different steps at a network edge node. In a group of user ports of a network edge node, data packets are received that include the same source address and the same destination address. Link aggregation of a group of user ports is performed by adding an internal header element representing each user port and the same network port to the received data packet at each user port, And is executed by a switching device that switches the received data packets from the user ports to the network port. At the network port, data packets are sent to the far-end network edge node to which the client device identified by the destination address is connected.

At the network port, the number of transmitted data packets is determined. Also, for each of the user ports, the number of each of the internally lost data packets during transmission from each user port to the network port is determined through the added internal header elements. Also, at a network port, a corrected number of transmitted data packets is determined using determined numbers of internally lost data packets during transmission from each user port to the network port. The modified number of transmitted data packets is then displayed on the far-end network edge node.
The modified number of transmitted data packets is the sum of the total number of internally lost data packets and the determined number of transmitted data packets during transmission from the user ports to the network port.

Figure 1 illustrates two network nodes and other data packet streams in accordance with one embodiment;
Figure 2 shows two edge network nodes and other data packet streams in accordance with a further embodiment;
3 illustrates different steps of a method for determining the total number of lost data packets during consecutive time intervals according to one embodiment.
4 shows different steps of a method for determining a modified number of received data packets;
5 illustrates a network edge node device according to one embodiment.
6 illustrates a network edge node device according to a further embodiment;
Figures 7A-7D show different steps of the proposed methods for determining a modified number of transmitted data packets.

The inventors have found that although packet processors located at network ports of two network edge nodes can determine the number of lost data packets between the network ports, But the actual number of lost data packets between the two clients may not be the same. Additional packet loss may be caused by loss of internal data packets within the network edge node. For example, a hardware failure of a sub-device of a network edge node may cause data packet loss. Also, data packets may be lost due to the fact that the packet processors of the network edge nodes store data packets in packet queues and may drop data packets for queue congestion reasons have. Therefore, internally lost data packets within a network edge node must be considered.

In addition, the client device may be connected to the network edge node through a plurality of user ports forming a link aggregation group as well as through a single single user port. Then, other sub-devices and other packet processors connecting the user ports of the link aggregation group and the network ports of the network edge node may cause data packet loss. Thus, when the protocol of packet loss measurement is executed between one packet processor located at the network processor of the network edge nodes and another packet processor located at the network port of the far-end network edge node, It will not take into account the internally occurring data packet loss between the node's network ports.

When operating separate protocols that measure packet loss in parallel between each packet processor located in one of the user ports and the packet processor located at the network port of the far-end network edge node, Lt; RTI ID = 0.0 > message exchange. ≪ / RTI > In addition, this may require coordination of different parallel protocols at the far-end network edge node. In addition, the far-end network edge node must consider the number of user ports forming the link aggregation group, which must provide separate protocol instance execution for each user port of the link aggregation group.

Wherein the number of data packets is a number transmitted from one client device to another client device connected to a far-end network edge node through user ports of a link aggregation group at a network edge node, The number of lost connections between the user ports and the network ports of the far-end network edge node, and the number should be determined for the purpose of performance monitoring.

It is an object of the present invention to provide a method of packet loss measurement by means of a direct protocol message exchange implementation between packet processors located at network ports of network edge nodes. Considering the internal data packet loss between the network port of the connected network node and the user ports of the link aggregation group, the client transmitting the data packets must consider the protocol executed at the network port of the far- No change is required. This is accomplished by the proposed method described in detail below.

1 shows a network edge node NEN1, which receives a plurality of data packet streams DPS1, DPS2 in a plurality of user ports UP1, UP2, representing a user network interface (UNI) do. The data packets of streams DPS1 and DPS2 include the same source address identifying the client transmitting the data packets and the same destination address identifying the client to which the data packets are sent. Each user port UP1, UP2 is arranged on one line card LC1, LC2. Each user port UP1, UP2 is arranged on a separate line card LC1, LC2. Also, a plurality of user ports may be arranged on the same line card.

Each line card LC1, LC2 includes a packet processor PP1, PP2 which processes the data packets of the received data packet stream DPS1, DPS2. The packet processors PP1 and PP2 provide packet queues Q1 and Q2 for temporarily storing data packets.

The line cards LC1 and LC2 are preferably connected to a switching device SD in the form of a single switch fabric or a plurality of switching cards. Further, the line card LC3 is connected to the switching device SD. The line card LC3 provides a network port NP1 representing the network-network interface (NNI) of the network edge node NEN1. Further, the line card LC3 includes a packet processor PP3. The packet processor PP3 provides a packet queue Q3 for temporarily storing data packets. According to the example shown in FIG. 1, a network edge node (NEN1) generally includes only one network port (NP1), while a network edge node may include a plurality of network ports in general.

The node NEN1 also includes a central control device (CCD) connected to the switching device SD and the line cards LC1, LC2, LC3 via the internal interface IIF. Through the interface IIF, the central control device (CCD) is connected to the switching device SD and the packet processor 12 to perform link aggregation of the user ports UP1 and UP2 to the link aggregation group LAG. (LC1, LC2, LC3).

Link aggregation of user ports (UP1, UP2) to the link aggregation group (LAG) is performed in the following manner. The packet processors UP1 and UP2 indicate to the respective received data packets of the streams DPS1 and DPS2 the user port on which the data packet was received and also the network port NP1 on which the data packet is to be switched Add an internal header element. In addition, switching device SD utilizes additional internal header elements of data packets to switch data packets to network port NP1. Resulting in a combined data stream (CDS) of data packets in combination with the data streams DPS1 and DPS2 and a point-to-point connection at the network port NP1. (P2PC) to the far-end network edge node (NEN2). A point-to-point connection is a connection between two systems, or alternatively between two network edge nodes, over a wide area network (WAN).

The packet processors PP1 and PP3 together determine the number of lost data packets during transmission of the data packet stream DPS1 from the user port UP1 to the network port NP1 in a manner to be described later. The packet processor PP1 receives from the client at the user port UP1 an internal header element which contains the same source address and the same destination address and identifies the user port UP1 and the network port NP1, And a counter for determining the number of data packets. The counter of the packet processor PP1 utilizes the received data packets before the received data packets are temporarily stored in the packet queue Q1. The packet processor PP3 provides a counter for determining the number of data packets received from the switching device SD and including an internal header element that identifies the user port UP1. The packet processor PP3 removes the inner header elements from the data packets. The data packets are then transmitted from the network port NP1 to the far-end network edge node NEN2. The counter of the packet processor PP3 utilizes the data packets before they are transmitted to the far-end network edge node NEN2 after reading the data packets from the packet queue Q3. The packet processor PP1 displays a determined number of data packets received in the packet processor PP3. This indication is performed by processor PPl which generates a data packet containing the number of data packets received. In order to obtain high measurement accuracy, the generated data packet is inserted after the last data packet which contributes to the determined number of received data packets. When the generated data packet is received, the packet processor PP3 subtracts the determined number of data packets received from the switching device SD from the number of received data packets indicated by the packet processor PP1, And determines the number of internally lost data packets during transmission from port UP1 to network port NP1.

The packet processors PP2 and PP3 together determine the number of lost data packets during transmission from the user port UP2 to the network port NP1 in a manner similar to that described above for the packet processors PP1 and PP3 do.

Also at the network port NP1, the packet processor PP3 is sent to the far-end network edge node NEN2 and determines the number of data packets including the same source and destination addresses.

The number of data packets received at the user network interface (UNI) from the client device identified by the source address and made to be transmitted to the client device connected to the far-end network edge node (NEN2) , The packet processor PP3 determines a modified number of transmitted data packets. The modified number of the transmitted data packets is

The number of transmitted data packets determined by the packet processor PP3 at the network port NP1, and

The sum of the total number of internally lost data packets between the user ports UP1, UP2 and the network port NP1.

 The total number of internally lost data packets is the sum of internally lost data packets in the network edge node NEN1 during internal transmission from user ports UP1, UP2 to network port NP1.

The network edge node NEN1 then indicates to the far-end network edge node NEN2 a modified number of data packets transmitted at the network port NP1.

Advantages of the proposed method are as follows: A packet processor located at the network port (NP2) of the far-end network edge node (NEN2) determines the number of received data packets, including the same source address and the same destination address, And to determine the actual number of lost data packets during transmission from the user ports UP1, UP2 to the network port NP2, The displayed modified number can be compared. To this end, the packet processor located at the network port NP2 must exchange protocol messages with the packet processor PP3 located at the network port NP1. The packet processor located at the network port NP2 does not utilize the number of the user ports UP1 and UP2 forming the link aggregation group LAG and the packet ports of the user ports UP1 and UP2 Lt; RTI ID = 0.0 > NP1) < / RTI > Therefore, the packet processor located at the network port NP2 can communicate with the network port NP2, as specified in Y.1731, to determine the actual number of lost data packets during transmission from the user ports UP1, UP2 to the network port NP2. The loss of data packets occurring between the user ports of the group (LAG) and the network port (NP1) can be compensated by the packet processor (PP3) located in the network port (NP1) do. This also means that the addition of a user port to a Link Aggregation Group (LAG) or the removal of a user port from a Link Aggregation Group (LAG) is not performed by a packet processor located at the network port (NP2) PP2, and PP3 of the NEN1 packet processors PP1, PP2, and PP3. In addition, the proposed method does not require implementation of a packet loss measurement protocol between each packet processor (PP1, PP2) in the user ports (UP1, UP2) and the packet processor located in the network port (NP2). Thus, the amount of protocol message exchange between nodes NEN1, NEN2 is not affected by the proposed method.

An indication of the determined actual number of lost data packets is performed by a packet processor (PP3) that generates a data packet that moves a data element representing a modified number of transmitted packets. The generated data packet is then transmitted to the network edge node (NEN2) via a point-to-point connection (P2PC). To obtain high measurement accuracy, a packet processor located at network port NP2 determines the number of data packets received at that time instance at which the generated data packet was received at network port NP2. The generated data packet itself does not consider the determination of the number of received data packets.

Further, the node NEN2 can display the number of data packets received at the network port NP2 at the node NEN1, and then, at the same time, between the user ports of the group LAG and the network port NP2 The actual number of lost data packets can be determined. The indication of the number of data packets received at the network port NP2 is sent to the packet processor located at the network port NP2 which generates a data packet that moves the data elements representing the number of data packets received at the network port NP2 Lt; / RTI > The generated data packet is then transmitted to the network edge node NEN1 via a point-to-point connection (P2PC). According to another alternative, the network edge node NEN2 may include a modified number of previously received and displayed data packets in the generated data packet for display on the network edge node NEN1, Corresponds to a modified number of transmitted data packets. This has the advantage that the network edge node NEN1 need not be associated with itself and that the modified number of transmitted packets corresponds to the number of received data packets.

Figure 2 shows the nodes NEN1, NEN2, which already have all the sub-devices, user ports and network ports described in accordance with Figure 1. 2 also shows a reverse combined data packet stream, which is transmitted by the network node NEN2 to the network node NEN1 through the network port NP2 and received at the network port NP1, 0.0 > (RCDS). ≪ / RTI > The data packets of the data stream RCDS include the same destination address as the source addresses of the data packets of the data streams DPS1 and DPS2 shown in Fig. In addition, the data packets of the data stream RCDS include the same source address as the destination address of the data packets of the data streams DPS1 and DPS2 shown in Fig.

The number of data packets lost among the above-mentioned clients should be measured in the direction of transmission opposite to the transmission direction used first in Fig.

At the network port NP1, the number of received data packets is determined by the packet processor PP3.

Link aggregation of a group of user ports UP1 and UP2 is also performed to transmit data packets from the node NEN1 to the user ports of the link aggregation group LAG and the connected client devices. For the received data packets, an inner header element is added which identifies one of the user ports (UP1, UP2) of the group (LAG). The user port identified by the inner header element may be computed by a load balancing principle known as a round robin or alternatively by computing a hash function through the data content of the data packet and then selecting the user port according to the resulting hash value. The switching device SD then switches the received data packets from the network port NP1 to the user port identified by the inner header element. As a result, the data stream RCDS is divided into parts of the data streams RCDS1 and RCDS2. At user ports UP1 and UP2, data packets are sent to the client device.

How many data packets internally lost during transmission from the network port NP1 to the respective user ports UP1 and UP2 are determined for each user port UP1 and UP2 of the link aggregation group LAG. For the user port UP1, the determination is performed by the packet processor PP1 together with the packet processor PP3. For the user port UP2, the determination is performed by the packet processor PP2 together with the packet processor PP3.

Packet processor PP3 provides a counter which is received at network port NP1 and which determines the number of its data packets belonging to the data stream RCDS. To these data packets, an internal header element is added which identifies one of the user ports UP1, UP2 of the group (LAG) and the network port NP1. The counter of the packet processor PP3 utilizes the received data packets before the received data packets are temporarily stored in the packet queue Q3. The processor PP3 also provides a counter for determining the number of data packets assigned to an internal header element that identifies the user port UP1 and the network port NP1.

The packet processor PP1 provides a counter for determining the number of data packets, including an internal header element, received from the switching device SD and identifying the user port UP1 and the network port NP1. The packet processor PP1 removes the inner header element from the data packets. The data packets are then transmitted from the user port UP1 to the client device. The counter of the packet processor PP1 utilizes the data packets before they are transmitted to the client device after the data packets are read from the packet queue Q1. The packet processor PP3 displays in the packet processor PP1 the number of data packets assigned to the inner header element that identifies the user port UP1 and the network port NP1. This indication is performed by the processor PP3 which generates a data packet containing the number. In order to obtain high measurement accuracy, the generated data packet is inserted after the last data packet which contributed to the determined number of data packets assigned to the user port UP1. The packet processor PP1 determines the number of data packets received in the time instance at which the generated data packet was received.

The packet processor PP1 itself is capable of transferring from the network port NP1 to the user port UP1 by subtracting a determined number of data packets received from the switching device SD from the number of data packets indicated by the packet processor PP3 And determines the number of internally lost data packets during transmission. The determined number of internally lost data packets then generates another data packet containing a determined number of internally lost data packets, and further from the user port UP1 to the network port < RTI ID = 0.0 > Is displayed on the processor PP3 by the processor PP1 by inserting the data packet into a packet transmission directed to the processor NP1. The packet processor PP1 also generates a data packet containing a determined number of data packets received at the port UP1 from the switching device SD and a displayed number of data packets previously indicated by the packet processor PP3 And can be inserted. Upon receiving the data packet, the processor PP3 may determine the number of internally lost data packets during transmission from the network port NP1 to the user port UP1.

 In a similar manner as described above for the packet processors PPl and PP3, the packet processors PP2 and PP3 together with the number of lost data packets during transmission from the network port NP1 to the user port UP2 .

At network port NP1, the modified number of received data packets is internally lost during transmission of data packets from network port NP1 to user ports UP1, UP2 from a first determined number of received data packets Is determined by subtracting the total number of data packets. The total number of internally lost data packets is determined as the sum of the lost data packets during transmission from the network port NP1 to the respective user ports UP1 and UP2 of the link aggregation group LAG.

The network processor located at the network port NP2 of the far-end network node NEN2 includes the above-mentioned source address and destination address and is connected to the node NEN1 via the point-to-point connection P2PC at the network port NP2. Lt; RTI ID = 0.0 > NP1. ≪ / RTI >

The packet processor located at the network port NP2 uses the generated data packet to indicate the number of data packets transmitted to the network port NP1. The packet processor PP3 may then send the lost data packet < RTI ID = 0.0 > (PD2) < / RTI > during transmission from the network port NP2 to the user ports UP1 and UP2 by subtracting a modified number of received data packets from the displayed number of transmitted data packets Lt; / RTI > In addition, the processor PP3 uses the generated data packet to indicate a modified number of data packets received at the network port NP2. The packet processor located at the network port NP2 then extracts the user ports UP1 and UP2 from the network port NP2 by subtracting the indicated modified number of the received data packets from the determined number of transmitted data packets. Lt; / RTI > the actual number of data packets lost during transmission to the base station.

Figure 3 illustrates other steps in a method for determining the total number of internally lost data packets during transmission of data streams from user ports UP1, UP2 to network port NP1, in accordance with an alternative approach. For each of the user ports UP1 and UP2, the number of internally lost data packets between each user port UP1 and UP2 and the network port NP1 is determined in each periodicity. For the user port UP1, the number of internally lost data packets is determined in a period corresponding to a time period TP1. The number of internally lost data packets (ILDP11, ILDP12, ..., ILDP1N, ILDP1N + 1) for each time interval (TI11, TI12, ..., TI1N, TI1N + It is determined at the end of the interval. For the user port UP2, the number of internally lost data packets is determined in a period corresponding to the time interval TP2. The number of internally lost data packets (ILDP21, ILDP22, ..., ILDP2N, ILDP2N + 1) for each time interval (TI21, TI22, ..., TI2N, TI2N + Lt; / RTI >

The total number of internally lost data packets (ONLDP1) is determined in the whole cycle corresponding to the time interval TP3. For the total time interval TI31, the total number of internally lost data packets corresponds to the number of internally lost data packets (ILDP11, ILDP12, ..., ILDP1N, ILDP21) that were determined during the entire time interval TI31 , ILDP22, ..., ILDP2N).

The time intervals TP1 and TP2 need not be the same and may be different. This has the advantage that no exact synchronization is required between the other associated packet processors provided in the user ports UP1, UP2.

On the one hand, the determination of the number of internally lost data packets for each user port and, on the other hand, the determination of the total number of internally lost data packets, on the one hand, Lt; RTI ID = 0.0 > period. ≪ / RTI > In addition, the determination of internally lost data packets between the user port UP1 and the network port NP1 is made by determining the number of internally lost data packets between the user port UP2 and the network port NP1, May be performed on time instances that are not identical. Therefore, no synchronization is required between the user ports UP1, UP2 and the packet processors provided in the network port NP1, to enable the determination of the total number of internally lost data packets.

The number of lost data packets during the time interval TE1N + 1 is not processed within the total time interval TI31, but the number of lost data packets will be processed within the next total time interval TI32. Is applied to the number of data packets lost during the initial period of the time interval TI2N + 1, thereby contributing to the number of lost data packets (ILDP2N + 1) and thus also the loss determined during the entire time interval TI32 Lt; RTI ID = 0.0 > data packets. ≪ / RTI >

The total number of internally lost data packets for each successive time interval (TI31, TI32) then determines the respective modified number of transmitted data packets and also determines the number of each actual number of lost data packets Is used to determine. Each determined actual number of lost data packets is then accumulated and / or averaged over said successive time intervals (TI31, TI32) for the purpose of performance monitoring.

Referring to FIG. 3, a method for determining the total number of internally lost data packets between the user ports UP1, UP2 and the network port NP1 has been described and suggested. In connection with the present method, it is possible to internally communicate between the network port NP1 and each of the user ports UP1 and UP2 during the transmission of data packets from the network port NP1 to the user ports UP1 and UP2 in each period It is possible to determine the respective number of lost data packets in a similar manner for each user port. It is therefore also possible to determine the total number of internally lost data packets between the network port NP1 and the user ports UP1 and UP2 in the entire period. This is achieved without having to synchronize the packet processors provided in the user ports UP1, UP2 and the network port NP1. The total number of internally lost data packets between the network port NP1 and the user ports UP1 and UP2 may then be determined for consecutive total time intervals, Lt; / RTI > The actual numbers of data packets may then be accumulated and / or averaged over consecutive total time intervals for purposes of performance capability monitoring.

According to Fig. 4, steps of a method for determining a modified number of transmitted data packets according to another alternative are proposed. As described in detail above with respect to FIG. 3, for each of a plurality of consecutive total time intervals (TI31, TI32, TI33), each total number of internally lost data packets (ONLDP1, ONLDP2, ONLDP3) is determined . Within the steps S101, S111 and S121, each modified number (CNRDP1, CNRDP2, CNRDP3) of received data packets is based on the total number of each internally lost data packets (ONLDP1, ONLDP2, ONLDP3) . The modified number of received data packets (CNRDP1) is displayed on the far-end network edge node in step S102.

As described above, the data packets internally lost during the time interval TI1N + 1, shown in FIG. 3, are considered to be within the total number of internally lost data packets ONLDP1 of the entire time interval TI31 But is considered within the total number of internally lost data packets (ONLDP) of the total time interval TI32. Thus, theoretically, the resulting modified number of received data packets (CNRDP2) may be less than the modified number of transmitted data packets (CNRDP1). This indicates that the far-end network node has received a number (CNRDP1) indication as the number of data packets received at the network port (NP1) at the first time instance and a number (CNRDP2) as the number of data packets received at the network port Will be received at a later time instance, which means that the number is smaller than the number of previous time instances. Thus, in the far-end network node, the number of data packets received at the network port from one time instance to the next time instance will be marked as being actually decremented, and the modified number of received data packets will be from one time interval TI31 Are inversely related to each other in terms of the far-end network node when they represent a sequence of numbers of received data packets that gradually increase until another time interval TI32.

Therefore, it is ensured that the displayed modified number of received data packets does not decrease between successive time intervals TI31, TI32. It is checked in step S112 whether the modified number CNRDP2 of the data packets received during the current total time interval TI32 is smaller than the modified number CNRDP1 of the received data packets of the previous total time interval TI31 do.

If the modified number of received data packets (CNRDP2) is not less than the modified number of received data packets (CNRDP1), the method proceeds to step S113. In step S113, the modified number of received data packets (CNRDP2) is displayed on the far-end network edge node.

If the modified number of received data packets (CNRDP2) is less than the modified number of received data packets (CNRDP1), the method proceeds to step S113A. In step S113A, the modified number of received data packets (CNRDP2) is equal to or greater than the difference between the modified number of received data packets (CNRDP1) and the modified number of received data packets (CNRDP2) . Further, in the next step S113B, the modified number of received data packets is adjusted and displayed on the far-end network node.

In step S121, a modified number of received data packets (CNRDP3) of the entire time interval TI33 is determined. In step S121A, it is checked whether the modified number of received data packets (CNRDP2) is less than the modified number of received data packets (CNRDP1). In a small case, the modified number of received data packets (CNRDP3) is reduced in step S121B by the same number as the modified number of received data packets (CNRDP2) increased in step S113A. In addition, steps S122, S123, S123A, S123B are performed similar to steps S112, S113, S113A, S113B so that the displayed modified number of received packets never decreases between consecutive time intervals.

In addition, the modified number of data packets may represent the number of each of the data packets received during each full time interval (TI31, TI32). If the modified number of data packets received during the time interval TI 32 is negative as a result of processing the total number of internally lost data packets determined during one time interval TI 32, then the total number of internally lost data packets Is shifted to the next successive time interval (TI33) in consideration of said portion when determining a modified number of data packets received during the next successive time interval (TI33). This causes the displayed modified number of data packets to be non-negative and will be incompatible with the viewpoint of the far-end network edge node.

FIG. 5 shows a network node NEN1 with its sub-devices, as already described with reference to FIG. In the packet processor PP1 of the line card LC1, a transfer counter TX11 is provided. The transfer counters TX11 and TX21 count the number of each of the data packets received from the client device at each user port UP1 and UP2. In the packet processor PP3 of the line card LC3, a reception counter RX11 is provided and counts the number of data packets received at the network port NP1 from the user port UP1. The reception counter RX21 also counts the number of data packets provided at the packet processor PP3 and received at the network port NP1 from the user port UP2.

The transmission counter TX1 is provided in the packet processor PP3. The transmission counter TX1 counts the number of data packets transmitted from the network port NP1 to the far-end network edge node. The determination of the total number of internally lost data packets in the network node NEN1 is performed using sample values of counters TX11, TX21, RX11, RX21. In other words, the number of internally lost data packets is determined using transmit and receive counters that observe the number of data packets in a portion of the data streams DPS1, DPS2 received at the user ports UP1, UP2.

Before delivering the numbers, the packet processors PPl, PP2, PP3 provide respective packet queues Ql, Q2, Q3 to store the received data packets in the packet queues Ql, Q2, Q3. The transmission counters TX11 and TX21 and the reception counters RX11 and RX21 are installed such that any of the packet queues Q1, Q2 and Q3 is located between the counters TX11, TX21, RX11 and RX21. This has the effect that the counters TX11, TX21, RX21, RX11 can take into account the loss of data packets caused by any of the queues Q1, Q2, Q3.

At any time instance, the transmit counter TXl 1 is sampled by the packet processor PPl. Next, the packet processor PP1 generates a data packet with the sample value of the transfer counter, and inserts the generated data packet into the data stream DPS1.

In another time instance, the packet processor PP2 samples the value of the transmit counter TX21 and generates a data packet containing the sample value of the transmit counter TX21. The generated data packet is then inserted into the data stream DPS2.

In addition, the packet processors PP1 and PP2 add internal header information to data packets received and allowed at each user port UP1 and UP2, as described above.

In the packet processor PP3, the number of data packets received from the user port UP1 is counted by the reception counter RX11, utilizing internal header elements of the data packets. Upon reception at the packet processor PP3 of the data packet carrying the sample value of the transmission counter TX11, the reception counter RX11 is sampled. The generated data packets themselves are not considered by the receiving counter RX11. The number of internally lost data packets between the user port UP1 and the network port NP1 is then determined as the difference between the sample value of the transmission counter TX11 and the sample value of the reception counter RX11.

In a similar manner, packet processors PP2 and PP3 perform the determination of internally lost data packets between ports UP2 and NP1, which depend on the sample values of counters TX21 and RX21.

As a next step, the packet processor PP3 samples the value of the transfer counter TX1. The sample value of the transmission counter TX1 is used to indicate the number of data packets received via the user ports UP1 and UP2 of the link aggregation group LAG to the far-end network edge node, UP2 and the network port NP1 by the total number of internally lost data packets in the network edge node NEN1.

The total number of internally lost data packets is determined as the sum of the data packets lost between the other user ports UP1, UP2 and the network port NP1.

The modified number of data packets transmitted from the network node NEN1 to the far-end network edge node is determined as the sum of the sample value of the counter TX1 and the total number of internally lost data packets. The modified number is then displayed on the far-end network edge node at network port NP1.

In addition, the transfer counter itself may be initially incremented by the total number of internally lost data packets, and then sampled to obtain a sample value of the transfer counter TXl. The sample value is then displayed on the far-end network edge node through the network port NP1.

FIG. 6 shows a network edge node NEN1 with all the sub-elements and counters as already described with respect to FIG. Figure 6 also shows the receive counter RXl provided by the packet processor PP3 at the network port NP1. The receiving counter RX1 counts the number of data packets received at the network port NP1 from the far-end network edge node via the inverse-coupled data stream RCDS. The packet processor PP3 distributes the data packets of the inverse coupled data stream RCDS to the user ports UP1 and UP2 of the link aggregation group LAG as described above with reference to FIG.

In the packet processor PP3, the transmission counter TX12 is provided to count the number of data packets to be transmitted from the network port NP1 to the user port UP1. The packet processor PP3 also provides a transfer counter TX22 which counts the number of data packets to be transferred from the network port NP1 to the user port UP2. The transfer counters TX12 and TX22 analyze the internal header information of the data packet to determine whether or not they consider the data packet.

In the packet processor PPl, the receiving counter RX12 is provided to count the number of data packets received via the reverse data packet streams RCDSl. In the packet processor PP2, the receiving counter RX22 is provided to count the number of data packets received via the reverse data packet stream RCDS2.

The number of internally lost data packets during transmission from the network port NP1 to the user ports UP1 and UP2 using the sample values of the transmission counters TX12 and TX22 and the reception counters RX12 and RX22 is Is determined in a manner similar to the determination of internally lost data packets during transmission from the user ports UP1, UP2 to the network port NP1, as described above with reference to Fig.

At the network port NP1, the receiving counter RX1 is internally reduced by the total number of lost data packets. The total number of internally lost data packets is the sum of data packets lost during transmission of data packets from network port NP1 to user ports UP1 and UP2.

FIG. 7 illustrates other alternatives for determining a modified number of transmitted data packets.

Figure 7A shows a process PR1 that takes into account the number of lost data packets between a user port and a network port. The illustrated process PR1 refers to the user port UP1 and the transmission counters TX11 and RX11 shown first in FIG. Those skilled in the art will appreciate that the process PR1 shown in FIG. 7A with counters TX11 and RX11 can be performed on different user ports using corresponding transfer counters and receive counters It can be seen clearly.

According to Fig. 7A, the transmission counter TX11 is sampled in step S301. The sample value of the counter TX11 is also inserted into the generated data packet and then received by the packet processor which provides the receiving counter RX11. When a data packet carrying a sample value of the transmission counter TX11 is received, the reception counter RX11 is sampled in step S302. In the next step S303, the number of internally lost data packets (ILDP1) between the user port and the network port is determined as the difference between the sample values of the transmission counter TX11 and the reception counter RX11. In the next step S304, a counter (ONLDP) for counting the total number of lost data packets is internally increased by the aforementioned predetermined number of lost data packets ILDP1. After this step S304, the process PR1 returns to the step S301.

Performs a similar process to process PR1, which determines lost data packets between the other user ports and the network port of the link aggregation group, and also internally increases the counter (ONLDP) by the number of each of the lost data packets , The total number of internally lost data packets between all user ports and network ports of the link aggregation group can be counted in a counter (ONLDP).

7B shows a process PR2 for determining and displaying a modified number of transmitted data packets (CNTDP). In step S401, the counter ONLDP is sampled. In step S402, the transfer counter TX1 is sampled. In step S403, the modified number of transmitted data packets (CNTDP) is determined by increasing the sample value of the transmission counter TX1 by the total number of lost data packets (ONLDP). In the next step S404, the modified number of transmitted data packets (CNTP) is indicated on the network edge node through the network port.

According to the solution described by FIGS. 7A and 7B, the total number of lost data packets is determined by the counter ONLDP and continues to increase. As an alternative to the processes PR1 and PR2 shown in Figs. 7A and 7B, Figs. 7C and 7D show processes PR1 'and PR2'.

The process PR1 'for determining the total number of lost data packets (ONLDP) includes all the steps as described above in accordance with FIG. 7A. Process PR1 'also includes step S305 following step S304. In step S305, the reception counter RX11 is internally increased by the determined number of the lost data packets ILDP1. After step S305, the process PR1 'returns to step S301.

FIG. 7D shows a process PR2 'for determining a modified number of data packets transmitted according to an alternative scheme. The process PR2 'includes all the steps of the process PR2 already described above according to Fig. 7B. Process PR2 'also includes step S405, which follows step S404. In this step S405, the transmission counter TX1 is incremented by the total number of data packets (ONLDP) lost. At next step S406, the counter ONLDP, which counts the total number of lost data packets, is reset to zero. After step S406, the process PR2 'returns to step S401.

Further, after sampling the counter ONLDP in step S401, the transmit counter TXl is immediately increased by the total number of ONLDP lost data packets, and the sample value is used as a modified number of transmitted data packets do. Also, the counter ONLDP is reset to zero.

The solutions described in accordance with Figures 7C and 7D are such that the counter (ONLDP) counting the total number of lost data packets is reset to zero. Accordingly, the process PR2 'can avoid the situation where the counter ONLDP increases infinitely. Therefore, an alternative solution according to FIGS. 7C and 7D is that the counter (ONLDP) can be implemented with a finite number of bits representing a counter (ONLDP).

According to a preferred embodiment, the packet processors are provided in the form of FPGAs (Field Programmable Gate Arrays). The line cards and the switching device are connected via a backpanel traffic interface.

The functions of the various elements shown in FIGS. 1, 2, 5 and 6, including all functional blocks designated as "processors ", are intended to encompass not only dedicated hardware, but also software Can be provided through the use of existing hardware. When provided by a processor, the functions may be provided by one dedicated processor, one shared processor, or a plurality of respective processors, some of which may be shared. In addition, the term " processor "or" controller "should not be construed to refer to only hardware capable of executing software, programmable gate arrays, read only memory (ROM) for software storage, random access memory (RAM), and non volatile storage. Other conventional and / or conventional hardware may also be included. Likewise, all the switches shown in Figures 1, 2, 5 and 6 are conceptual only. Their functions can be performed through the use of program logic, through dedicated logic, through the interaction of program control and dedicated logic, or passively, and in this context, Likewise, specific techniques may be selected by the implementer.

It will be appreciated by those skilled in the art that all the block diagrams herein depict a conceptual diagram illustrating a circuit diagram that implements the principles of the present invention. Likewise, all flow charts, flow diagrams, state transition diagrams, pseudo code, etc., represent various processes that can be substantially represented in a computer-readable medium, And may be executed by something other than a computer or processor as explicitly indicated.

Claims (10)

CLAIMS What is claimed is: 1. A method for enabling packet loss measurement in a network,
- receiving data packets in the group (LAG) of the user ports (UP1, UP2) including the same source address and the same destination address,
Link aggregation of the group (LAG) of the user ports (UP1, UP2)
An internal header element indicating each of the user ports UP1 and UP2 and the same network port NP1 in each of the user ports UP1 and UP2 of the group LAG, , ≪ / RTI &
Performing at the switching device SD by switching the received data packets from the user ports UP1, UP2 to the same network port NP1 using the added internal header elements;
- at said network port NP1,
Sending the data packets to the far-end network edge node NEN2,
- determining the number of transmitted data packets,
- for each of said user ports (UP1, UP2), internally lost data packets during transmission from said respective user port (UP1, UP2) to said network port (NP1) Determining a respective number,
- determining a modified number of data packets transmitted using said determined numbers of internally lost data packets during transmission from said respective user port (UP1, UP2) to said network port (NP1), characterized in that said transmitted Wherein the modified number of data packets is a sum of a total number of internally lost data packets and a determined number of the transmitted data packets during transmission from the user ports (UP1, UP2) to the network port (NP1) Determining a modified number of packets,
- displaying said modified number of data packets transmitted at said network port (NP1) to said far-end network edge node (NEN2). The method according to claim 1,
- at said network port NP1,
Receiving second data packets comprising a second source address that is the same as the first destination address and a second destination address that is the same as the first source address,
- determining a second number of said second data packets received,
- link aggregation of the user ports (UP1, UP2)
At said network port (NP1), an internal header element indicating one of said user ports (UP1, UP2) is added to each of said received second data packets,
Performing at the switching device SD by switching the received second data packets to one of the user ports UP1, UP2 using the added internal header elements;
- determining the number of each of the internally lost second data packets during transmission from said network port (NP1) to said respective user port (UP1, UP2) for each of said user ports (UP1, UP2) ,
- determining a modified number of received second data packets using said determined numbers of internally lost second data packets during transmission from said network port (NP1) to said respective user port (UP1, UP2) ≪ / RTI > further comprising the steps of: The method according to claim 1,
Each of said internally lost data packets during transmission from said respective user port (UP1, UP2) to said network port (NP1) in each periodicity for each of said user ports (UP1, UP2) (ILDP11, ILDP21, ..., ILDP1N, ILDP2N)
The total number of internally lost data packets during transmission from the user ports (UP1, UP2) to the network port (NP1) in the entire period over the entire time interval (TI31) corresponding to the whole period, (ILDP11, ILDP21, ..., ILDP1N, ILDP2N) of internally lost data packets during transmission from the user ports (UP1, UP2) to the network port (NP1) Gt; wherein the method further comprises determining a packet loss measurement in the network. 3. The method of claim 2,
- each said number of internally lost second data packets during transmission from said network port (NP1) to said respective user port (UP1, UP2) for each of said user ports (UP1, UP2) ; Determining
(ONLDP1) of internally lost second data packets during transmission from said network port (NP1) to said user ports (UP1, UP2) over the entire time interval corresponding to said whole period Further comprising determining, by summing the determined respective numbers of lost second data packets during transmission from the network port NP1 to the respective user ports UP1 and UP2, How to enable it. 5. The method of claim 4,
The total number of internally lost second data packets (ONLDP1, ONLDP2) is determined during consecutive total time intervals (TI2, TI3)
At least a portion of the total number of internally lost second data packets (ONLDP2) determined during the first time interval TI2 is greater than or equal to at least a portion of the second total time interval TI3 Gt; (CNRDP3) < / RTI > of the first and second data packets in the network. The method according to claim 1,
At the network port NP1,
- receiving an indication of the number of data packets, including the source address and the destination address, received by the far-end network edge node (NEN2)
- the number of data packets lost during transmission from said user ports (UP1, UP2) to said far-end network edge node (NEN2), by said far-end network edge node (NEN2) from said modified number of transmitted data packets Further comprising determining by subtracting the indicated number of received data packets. The method according to claim 1,
In the far-end network edge node NEN2,
- receiving an indication of the modified number of the transmitted data packets;
- determining a received number of data packets comprising the source address and the destination address,
- subtracting the determined number of packets from the modified number of transmitted data packets by the number of lost data packets during transmission from the user ports (UP1, UP2) to the far-end network edge node (NEN2) Further comprising the step of determining a packet loss in the network. 3. The method of claim 2,
- displaying, at the network port (NP1), the modified number of the received second data packets at the far-end network edge node (NEN2)
- at said far-end network edge node (NEN2)
Determine the number of second data packets transmitted,
- receiving an indication of the modified number of the received second data packets,
The number of second data packets lost during transmission from said far-end network edge node (NEN2) to said user ports (UP1, UP2), from said determined number of second data packets, Further comprising the step of determining the number of packets to be transmitted by subtracting the modified number indicated. 3. The method of claim 2,
At the network port NP1,
- receiving an indication of the number of second data packets transmitted by the far-end network edge node (NEN2)
- the number of lost second data packets during transmission from said far-end network edge node (NEN2) to said user ports (UP1, UP2) from the indicated number of second data packets, ≪ / RTI > wherein the method further comprises: determining a packet loss measurement in the network. - at least one first line card (LC1) comprising at least one group of user ports (UP1, UP2) and at least one packet processor (PP1, PP2) connected to said user ports (UP1, UP2) Wow,
- at least one second line card (LC3) comprising at least one network port (NP1), and a packet processor (PP3) connected to the network port (NP1)
- a network edge node comprising a switching device (SD) connecting said first and said second line cards (LC1, LC2, LC3)
Wherein the first line card LC1 is configured to receive data packets through the user ports UP1 and UP3 and the second line card LC3 is configured to receive data packets via the network port NP1, Is configured to transmit data packets to a node (NEN2)
The packet processors (PP1, PP2) of the first line cards (LC1, LC2) are connected to data packets received via the user ports (UP1, UP2) and containing the same source address and the same destination address, To add an internal header element indicating the respective user port (UP1, UP2) and the network port (NP1) from which the data packet was received,
The switching device SD may be configured to switch the received data packets from the user ports UP1, UP2 to the network port NP1 using the added internal header elements,
Wherein the packet processor (PP3) of the second line card (LC3) is configured to determine the number of data packets sent to the far-end network edge node (NEN2) and comprising the same source address and the same destination address,
The packet processors PP1, PP2 and PP3 transmit the network port NP1 from the respective user ports UP1 and UP2 to the user ports UP1 and UP2 using the added internal header elements, To determine a number of each of the internally lost data packets during transmission to the base station,
The packet processor (PP3) of the second line card (LC3)
- to determine a modified number of said transmitted data packets using said determined number of internally lost data packets during transmission from said respective user port (UP1, UP2) to said network port (NP1) The modified number of data packets being internally lost during transmission from the user ports (UP1, UP2) to the network port (NP1) is the sum of the total number of data packets lost and the determined number of the transmitted data packets,
- display said modified number of transmitted data packets on said far-end network edge node (NEN2).

Images